The primary goal of this proposal is to define in molecular terms the conformational changes that underlie C- type inactivation in K+ channels, and explain the implications of these conformational changes with regards to ion selectivity. C-type inactivation of K+ channels is a molecular process of great physiological significance and understanding the molecular basis of inactivation has a direct impact on human health. This goal will be achieved using a combination of computational and experimental approaches via 3 specific aims. In Aim 1, we will test the hypothesis that the constricted filter conformation, as observed in several crystal structures of the KcsA channel, is a general property of different K+ channels. Computations are the best way to address this issue to examine three channels that can inactivate and or not inactivate, as well as several mutants of these channels. The impetus to carry this work comes from our recent discovery of the role of inactivating water molecules to explain the origin of the multi-second timescale for recovery in the KcsA channel (Ostmeyer et al, Nature 2013). In Aim 2, we will test the hypothesis that multi-ion effects can explain the difference in ion conduction and selectivity using multi-ion potential of mean force (PMF) computations. In particular, we will test whether the height of the multi-ion free energy barrier through the non-conductive (constricted) state explains the wide range of behaviors in permeation and selectivity that are observed. To buttress our analysis of constricted pores, we will also exploit recent high-resolution data to test the mechanism of selectivity through conductive filters using multi-ion 3d-PMFs. In Aim 3, we will test the hypothetical concept of pore dilation and discover alternative filter conformations using the Tsuka channel as a model system. We recently determined the crystal structure of the Tsuka channel, which displays a novel filter conformation that is wider than usual. Tsuka has about 30% similarity to the NaK channel, which can be converted into a KcsA-like conductive pore with a few mutations. This implies that Tsuka is a relevant member of the K+ channel family. This crystal structure of a novel channel of relevance provides the impetus for this work. In MD simulation, this open dilated pore does not conduct. We will exploit this novel crystal structure to further test the concept of pore dilation and, using MD simulations, examine whether it can behave as a non-conductive inactivated filter. We will also progressively re-engineer the sequence of Tsuka to recapitulate the properties of a K+ selective pore, with the intention of capturing conformational intermediates spanning the range from dilated to conductive selective pore. This sequence/structure transformation will be mapped using X-ray crystallography, electrophysiology, MD simulations, and ROSETTA-DESIGN computer-assisted protein engineering.